Prediction of vertebral fractures in patients with monolateral adrenal incidentalomas

Context: Subtle hypercortisolism is associated with an increased risk of vertebral fracture (VFx). Objective: The objective was to determine the best parameters of cortisol secretion for detecting the VFx risk in patients with adrenal incidentalomas (AI). Design: This was a retrospective (cross-sectional arm) and prospective (longitudinal arm) design. In the cross-sectional arm, we assessed the accuracy of the cortisol secretion indexes in identifying the patients with VFx (prevalent VFx). In the longitudinal arm, we tested the cortisol secretion parameters, which were able to identify the prevalent VFx, for the prediction of the occurrence of a new VFx (incident VFx) in AI patients followed-up for at least 2 years. Setting: Four referral Italian endocrinology units participated in this study. Patients: A total of 444 and 126 AI patients without symptoms of hypercortisolism enrolled in the cross-sectional arm and longitudinal arm, respectively. Main Outcome Measures: Serum cortisol after a 1-mg dexamethasone suppression test (1 mg DST), urinary free cortisol, ACTH, bone mineral density at lumbar spine and femoral neck (by dual-energy X-ray absorptiometry), and the VFx presence (by X-ray). Results: The cortisol levels after 1 mg DST that were greater than 2.0-g/dl (55 nmol/liter) were the best criteria for detecting patients with both prevalent (73.6% sensitivity, 70.5% specificity) and incident VFx (80% sensitivity, 68.8% specificity) and were associated with a 10-fold increased risk of a new VFx (odds ratio,10.27; 95% confidence interval, 3.39-31.12; P <.0001), regardless of age, gender, bone mineral density at lumbar spine, and prevalent VFx. Conclusions: In AI patients without symptoms of overt hypercortisolism, the cortisol levels after 1 mg DST greater than 2.0-g/dl (55 nmol/liter) represent the best criterion for detecting prevalent and incident VFx.